Pr edictive Maintenance of Induction Motor s

With MCSA & TSA Technology

R K Gupta
Pr incipal Consulting Electr ical Engineer (Power System)
Reliability Engineer
BE Electr ical
Pr otection, A utomation, PLC & SCA DA , Level 3 MCSA , Colorado
USA
What is MCSA & TSA ?

MCSA : stand for Motor Current Signature Analysis. The basis of MCSA is that the
stator current contains current components which is directly linked to rotating
magnetic flux components caused by electrical or mechanicals faults. These
harmonics current components caused by fault can be used for early failure detection.

TORQUE SIGNA TURE: The acquisition of torque ripple and the analysis of the torque
spectrum is a relatively new addition in the realm of condition based maintenance
technology. By capturing the real components of the current and deriving the flux
generated from voltage a signal equal to air gap torque is developed. Mechanical
information buried in a complex modulated current signal can now be easily accessed
and interpreted for electrical and mechanical conditions.
Amazing Fact

World’s 30% - 40% produced electrical energy is consumed by electrical

motors.
 Dynamic Motor Testing

 A motor is only one part of a “Machine System” that includes: Power in, the Motor
and the Load.
 Dynamic testing diagnoses issues for all three links .
 Gives clarity to the Root Cause Analysis
 Power Quality, Motor, Load issues
 Helps to find
 Mechanical issues including: Bearing problems, Rotor Bar issues, Cavitation's
and Eccentricities
 Electrical issues including: Oscillations, Transients and Harmonics (Distortion)
Benefits of Dynamic testing

 Helps separate Mechanical from Electrical issues

 Provides “early warning” information
 Provides “support” for other technologies
 Can be performed more often than static testing
 It is safe, fast and non-intrusive
Defining Dynamic Electric Motor
Monitoring

 WHAT IS IT?: The ability to find Power Condition, Motor and Load Related issues
while the motor is running under normal conditions.

 HOW?: We measure, analyze and trend the currents and voltages of the motor-
load system (typically at the MCC).
Maintenance Program Types

 Corrective – Run to failure

 Preventative – Pre-planned periodic motor replacement or recondition regardless

of condition, run time or history.

 Predictive – A complete program that includes static and dynamic motor testing,
etc. All diagnosed and trended.
 Cost Vs. Benefits

 Corrective $17 - $18 Per Hp

 Preventive $11 - $13 Per Hp
 Predictive $ 7 - $ 9 Per Hp

 1983 EPRI Study of Maintenance Programs

 Motor Failure Areas

IEEE (early 1990’s) EPRI (Mid 1990’s)

Bearing Bearing
44% 41%

Other Stator Other

Rotor Rotor Stator
22% 26% 14%
8% 9% 36%
What are we really after?

 Reduce unscheduled downtime by predicting imminent motor failures and

identifying problem areas
 Determine root-cause of problem
 Ultimate goal: Save time & money
The Questions:

 Is the motor operating properly?

 Is the load operating properly?
 How is the motor reacting to the load?
 What is the condition of the power?
To Find the Answers:

 Observe voltages & currents supplying motor.

 Obtain motor speed.
 Obtain motor torque.
 Analyze data.
 Trend routinely.
How it Works ?
Chain of Events

1. frequency 2. speed
Load

Motor

3. Torque
MCC 4. Power

5. Voltage

6. Current
 Action & Reaction

Here is how this thing works: The MCC tells the motor: ‘I am sending you 50Hz.’
Motor replies ‘Ok. Load, I would like to turn you at this speed ‘(1500rpm, say). Load
says: ‘Ok, you want to do that, then give me this much torque’. Motor responds ‘Fine’.
Turns around and says to the MCC: ‘I need this much power’ (remember, mechanical
output power of the motor is speed times torque. Both of those parameters are set
right now). MCC says ‘Well, I know that I am giving you this particular voltage here’.
Then the motor says ‘All right, granted that you offer me this voltage, then give me this
amount and type of current’.
In short: If you find anything wrong with the torque, it m us t be caused by the load. If
you find anything wrong with the voltage, it m us t be caused by the MCC. If you find
anything wrong with the current, then it is the play of MCC, motor and load together.
 Note

 Torque related problems are caused by the load

 Power related problems come from “upstream” and effect the whole buss.
 If you find anything wrong with the current then it is the play of the MCC, motor and
load together.
Safety
Load
Breaker

Motor

MCC
Safety First!
Explorer
NFPA 70
“Best Practices”
Your company’s requirements
Your industry’s requirements
Connection

V
oltageProbes

Current CTs
Connection for LT/LV Motors

VoltageProbes

Current CTs
Connection for HV/HT Motors

Load
Breaker

Motor

CTs

PTs
 No physical connection to motor required.
 No transducers required to be installed on the motor
 Permanent installation available
 All values derived from voltage and current measurements only

Explorer
Connection with EP

Load
Breaker

Motor

CTs
PTs 1 of 700+ EPs at
Entergy
EP

Explorer First Energy

RC Pump
Speed – accuracy is critical
 Accuracy of Speed Critical:

Consider a 4 pole motor:

Synchronous speed = 1800RPM
Full Load Speed = 1780 RPM
 1% Error in speed measurement = 18 RPM!
 Must be able to resolve shaft speed to the 1RPM level or better: 0.05% (0.05% =
1/1800*100%)
 Special signal processing techniques required: DFLL (Digital Frequency Locked
Loop) – basically a form of correlation.
 FFT will not work due to very long observation times (60 seconds or more) to get
enough frequency resolution to resolve 1 RPM.
Current Signature: FFT vs. DFLL

FFT DFLL
Amplitude: 60dB

Amplitude: 20dB

Resolution: 0.13Hz Resolution: 0.005Hz

FFT vs. DFLL
Determination of Motor Speed
Motor Speed

 Air gap changes as rotor rotates within the stator core because of excentricities,
misalignments of rotor and stator, out of round rotors, etc.
 Current in stator changes (slightly) as a result of the changing gap distance.
 Change in current shows up as a peak in the current spectrum.
 Torque Derivation

 Use “dq0” theory – also called two axis theory.

 Theory exists since 1929
 R.H. Park, “Two-Reaction Theory of Synchronous Machines – Generalized Method of
Analysis. Part 1”, AIEEE Transactions, Vol. 48, July 1929, pp. 717-717.
 Used in VFD’s since the 1980s
 E.C. Lee, “Review of Variable Speed Drive Technology”, www.powertecmotors.com/avsde4.pdf.
 T.A. Lipo, A.B. Plunkett, “A Novel Approach to Induction Motor Transfer Functions,” IEEE
Transactions on Power Apparatus and Systems. Vol. PAS 93 pp. 1420-1419, 1979.
 A. B. Plunkett, “A Current Controlled PWN Transistor Inverter Drive,” IEEE/IAS 1979
Annual Meeting, pp 785-792.
 Well documented in motor control texts.
 P.C. Krause, O. Wasynczuk, S.D Sudhoff, “Analysis of Electric Machinery,” IEEE Press
NY, ISBN 0-7803-1101-9, 1995.
 Torque Shows What the Load is
Doing

 Variations in torque indicates a problem

with the smooth operation of the motor
and or load.
 Intermittent load variations -> T(t)
 Repetitive load variations -> T(freq)
 To the right: Fan with a flapping belt
causing excessive bearing wear on both
motor and fan pillow blocks.
Overall Check

 Power Quality
 Voltage level, voltage unbalance, harmonics distortion, total distortion, power, harmonics
 Machine Performance
 Payback period, effective service factor, load, operating condition, efficiency
 Current
 Current level, current unbalance
 Spectrum
 Rotor bar, V/I spectra, demodulated spectra
 Connections
 Waveforms, ABC/SYM components, phasors
 VFD details
 Torque and speed vs. time, frequency and voltage vs. time.
Summary page – After each test
 Power Quality

 Power level
 Voltage unbalance
 Harmonic distortion
 Total distortion
 Power
 Harmonics
 Voltage Level

Present
test
Trending
Log
Percent
Under
Voltage
Incoming Power - Voltage Issues

Low Voltage Over Currents (Over Heat)

High Voltage Low Power Factor

Iron Saturation

Ultimately Higher Losses

 Over/Under Voltage

 Motors are designed to operate with +/- 10% of rated voltage

 Ideally, voltage supply deviation should be less than +/- 2%
 When operating over/under voltage a motors performance, efficiency, and power
factor change
 Over/Under Voltage

 Voltage deviations usually caused by

 Poorly performing or improperly adjusted transformers
 Undersized conductors
 Poor connections
 Low power factor sources in the distribution system
Acceptable System Voltage Ranges

Nominal System Allowable Limits % Allowable Voltage

Voltage Range

120V (L- N) +/
-5% 114V- 126V

240V (L- L) +/
-5% 228V- 252V

480V (L- L) +/
-5% 456V- 504V
Comparison of voltage level to average
winding temp & motor efficiency*

Voltage -10% Normal +10%

(414V) (460V) (506V)
HP Full
Load Temp Ef
f Temp Ef
f Temp Ef
f

10 66 90.0 56 91.4 55 91.5

20 84 90.4 70 91.8 67 92.1

50 84 91.9 69 93.1 62 93.6

100 82 94.2 72 94.8 69 94.9

200 90 94.9 77 95.5 74 95.7

Over Voltage – Case study
6.6kV 1000 hp Pump Motor at a
Nuclear Power Plant

 Motor being overdriven. Operating at 7.7kV

 Plant engineering thought there should be no problem.
 Currents are below nameplate – so its ok, right?
 Wrong! Saturating the iron in the stator caused overheating
 Motor has burned up 3 times - $150,000 to repair.
 Thermal output of plant reduced to 50% during 3-4 weeks per repair.
 Lost $10,000,000 in electrical sales during each repair.
 Voltage Unbalance

Trending
Percent Log
Of
Unbalance
 Voltage Unbalances

 When a voltage unbalance reaches 5 %, the phase currents can differ by as much
as 40 %.

Unbal = 100 Vmaxdev-V

V
Where:
Unbalance = Voltage unbalance in %
Vmaxdev = Line to line phase voltage deviating most
from mean of 3 phases
V = RMS voltage, mean line to line of 3 phases
Effects of Voltage Unbalance on
Motor Losses
 Voltage Distortion

Trending
Log

Percent of
Distortion
Voltage distortion: Torque Spectra

Normal Peaks

60Hz
120Hz
360Hz

Clear – no distortion
Harmful distortion
Voltage Distortion – Case study
Harmonics Can Damage Motors

 Shotton Paper (Flintshire, UK, Shaw Bowden, shaw.bowden@upm-kymmene.com)

 Large inverter fed motor on the same bus as a smaller induction motor.
 Harmonics from the VFD destroying smaller 660V, 550hp motor.
 Voltage THD = 11%, 9%, 8% (three phases)
 Voltage Imbalance = 3.6%
 NEMA de-rating = 60%
 Running at Effective Service Factor = 1.29
 End result, motor running very hot.
 Time to get out the Baker Advanced Winding Analyzer to check for insulation
damage.
Harmonics Can Damage Motors
Power

T
ot
al
KwU
se
d
P
owe
rFac
to
r
A
ve
ra
geVol
t
age
A
ve
ra
geCur
re
nt
D
is
t
or
tio
ninV
ol
t
age
D
is
t
or
tio
ninC
ur
re
nt
V
ol
t
ageUnb
al
an
ce
C
ur
re
ntUnb
al
an
ce
Harmonics
Harmonics
 Machine performance

 Effective Service Factor

 Load
 Operating Condition
 Efficiency
 Payback Period
 Eff. s.f. = % Load
% NEMA derating
Test station 300 hp 3570 rpm
Effective Service Factor

Load 92%

Power Quality Good

“Motor Healthy”
Effective Service Factor

M
ot
orO
K
D
at
ain d
icates t
ha
tth
is
mo
to
r isprob ab
lyn
ot
o
ve
rheatin
g .
L
oadiso v
er 100%bu
t
p
owe
r q
u a
lityisg
ood
Load History

Percent Trending
Load Log
 Motor Performance
Percent Load vs Temperature

Tempe
ratur
e(C)
Hor
sepowe
rFul
lLoa
d 1.15% 1.25%
10 49 64 77
20 56 75 91
50 75 102 128
100 64 80 94
200 69 89 106
* Courtesy U S Motors
Operating Condition

Percent Trending
Change Log
 Operating Condition

 Goal: To compare machine’s operation with the past.

 Method:
 If past data exists at present torque & frequency level, then
 Is the voltage level similar to previous?
 Is the current level similar to previous?
 Is the phase angle similar to previous?
 Successes
 Flagged very early rotor bar deterioration
 Flagged high resistance connection on terminal box
 Issue
 Undiagnosed flags DO happen
Efficiency vs Load Level

Percent
Efficiency

Percent
Load
Efficiency

Percent Load
oadvsLosses inKw Percent LoadvsEfficiency
Money Savings/Efficiency Matters!

Pout
Efficiency η=
Pin Input Power

Output Power = Shaft Power

Input Power = Electr ical Power
Fundamental Relationships on
Efficiency:

Efficiency = 100% - Losses[%]

Efficiency = Output Power / Input Power

Example: 100hp motor at 90% efficiency,

with per fect voltage condition.

a) What is the Input Power and the Output Power ?

Pout = 100hp
Pin = 100hp / 0.9 = 111hp

b) What are the losses? Ploss = 111hp – 100hp

Ploss = 11hp = rated losses
Fundamental Relationships on
Efficiency:

Example: 100hp motor at 90% efficiency,

with 2% Voltage unbalance.

c) What is the Input Power and the Output

Power?
Pout = 100hp

Ploss = 11hp * 1.1 = 12.1hp

Pin = Pout + Ploss = 112.1hp

Testing Motors*

*LTEE
Variability Study of Identical
Motors

Motor HP No. of Test Efficiency Average Variation

Max. Min. Ave. Losses Losses (1)
200 10 96 .6 9 6.0 96.3 5.722 +7.9%

75 6 93.9 92.8 93.2 4.06 +6.5%

40 7 91.0 90.0 90.6 3.085 +7.9%

(1) Maximum Percent increase in Total Loss from the Average Loss
Payback Period

H
our
sp
erD
ay D
ay
spe
rWe
ek
C
os
tp
erK
wH
Payback Period

A
t97
.3
%Lo
ad

9
5.6
%Ef
f
ic
ie
ncy

A
n n
ual
Co stt
o
o
p e
ra
tet
h is
m o
to
rat
t
hisl
oa
dle v
el
$92,927
.84
 Current

 Over current
 Current unbalance
 Over Current

Percent Trending
Of Log
Current
 Current Unbalance

Percent Trending
Of Log
Unbalance
 Current Issues

 Over Current
 Increased heat causing faster degradation in the insulation system

 Current Unbalance
 Can be caused by voltage unbalance
 Can indicate internal motor problems – or
 Connection problems
 Spectrum

 Rotor Bar
 V/I Spectrum
 Demodulated Spectrum
 Harmonics
Rotor Bar
 Symptoms
 Symptoms of Broken/Degradation
Rotor Bar s
 High Rotor Impedance
 Increase in Rotor Current
 Heat/Temperature
 Effect of Br oken Bar s
 Requires constant torque level or it will
pulsate
 Next one breaks sooner
 Current increases
 Temperature increases
 Insulation life shortens
 Typically non-immediate death
Known good rotor bar

Signature of a
Good Rotor Bar
Broken Rotor Bars

Vertical Line
Identifies Rotor Bar
Frequency
Fan 1 hp 1740 r pm
Rotor Bar

MCSA detects current components induced in the stator winding due to rotating flux
components caused by problems such as broken rotor bars or high air gap eccentricity. It
is true that broken rotor bars will result in a change to the vibration spectrum, but vibration
is traditionally sensed at the bearings and for each motor there is a different mechanical
stiffness between the electromagnetic forces caused by broken bars or air gap eccentricity
and the position where the vibration is sensed. Consequently the vibration from
electromagnetic forces is a second order effect compared to current components directly
induced from the specific rotating flux waves. This is not the case with MCSA. With
respect to detecting air gap eccentricity problems a similar reasoning applies. It has also
been shown that the current spectrum changes due to mechanical misalignment and due
to abnormal rotor drive dynamics.
DETECTION OF BROKEN BARS

 Pr imar y Causes:

 Dir ect on-line (DOL)

 The starting duty cycles which the rotor cage winding was not designed to withstand.
This causes high thermal and mechanical stresses since the starting currents are
typically of the order of five times the full-load operating current.

 Pulsating Mechanical Load

 Pulsating Load arise when driving reciprocating compressors, ESP, Agitating
Pumps etc which can subject the rotor cage to high mechanical stresses.
DETECTION OF BROKEN BARS

Broken rotor bars result in current components being induced in the stator winding at
frequencies given by Equation

fsb = f1(1 ± 2s) Hz

CURRENT SIGNATURE DUE TO
BROKEN BARS

f1 =SUPPLY FREQUENCY (50 Hz)

LOWE SIDEBAND UPPER SIDEBAND

dB

Hz
SIGNATURE PATTERN DUE TO BROKEN BARS
TWICE SLIP FREQUENCY SIDEBANDS

dB
f1
-2sf1 +2sf1
fsb = f1±2sf1

Hz

IDEALISED Current Spectrum – Due to Broken Rotor Bars

 Case Study

 Vedanta A luminum Extr action Plant,

India
 Motor Details
 ID Fan
 V=6600
 I=40 A
 HP=475.87
 Speed=992
Air Gap Eccentricity

 Air Gap Eccentricity consists of two types :

 Static
 Dynamic

Both air gap exists simultaneously in a r eal SCIM due to manufactur ing
toler ances & installation pr ocedur e ( in lar ge motor s).
 Static Eccentricity

Static eccentricity where the position of minimum radial air gap length is fixed in space.
Causes of this include:

 Manufacturing tolerances on each component part.

 Stator core ovality.
 Out of tolerance spigots in the end frames that house the bearings.
 Incorrect installation of large HV motors.
 Dynamic Eccentricity

 Dynamic eccentricity where the minimum air gap revolves with the rotor and is a
function of space and time.
 Causes include –
 Non-concentric rotor and shaft.
 Thermal bow of the rotor.
 Combination of high air gap eccentricity, consequential unbalanced
 magnetic pull (UMP), and a flexible rotor.
 Severe bearing wear.
 Rotor core lamination movement independent of the shaft.
Dynamic Air Gap Eccentricity

Minimum air gap revolves with the rotor and it is the function
of space & time.
Diagnosis of Air Gap

Motor Current Signature Analysis (MCSA) is used to identify the characteristic current
signature pattern which is indicative of abnormal levels of air gap eccentricity. The
current signature pattern is identified via the following equation

fec = f1(R(1-s)/p ± nws) ± f1(1-s)/p

Consists of two par ts:
fr s = nor mal r otor slot passing fr equencies = f1(R(1-s)/p ±
nws)
fr = r otor speed fr equency = f1(1-s)/p = Nr (r pm)/60
Air Gap Eccentricity Graph

High Air Gap Eccentricity Normal Air Gap Eccentricity

Case Study

Vedanta A luminum,
India
Presence of Dynamic eccentricity
Motor Details
Load : Pump
V : 6600
I : 83.17
Speed : 2991
HP : 1072.39
 V/I Spectrum

 Voltage
 Current
 Demodulated Spectra
Voltage Spectra

300Hz
180Hz

120Hz
Current Spectra

180Hz 300Hz

120Hz
Demodulated Spectrum

 Analyzes the dynamic behaviour of Motor Load System

 It is the demodulated spectrum of Torque Signature
 With Channel Control, the demodulated spectrum can be changed from torque to
current or voltage of any one phase
Demodulated Spectra
Torque Domain

Two tests are available in torque domain:

 Torque Ripple Displays the instantaneous torque requirements of the load

 Torque Spectra
 Torque Ripple

 Torque Ripple shows the ripple

vs rated torque

Rated Normal
Torque but High
Torque Ripple

A
ut
oFa
ct
or
ySa
nit
ar
yPu
m p
To
ro
nt
o,Ca
nad
a

C
avit
at
io
ns N
orm
al
Case Study

Vedanta A luminum, India

Motor Details
Function: Coal crusher , Gear
Attached
V=6600
I=34.70
Speed=981
HP=375.34
Torque Spectrum

Torque Spectrum Analysis helps to find issue related dynamic behaviour of motor
load system.

 Bearing Issues ( Motor DE, Load End)

 Analyze Load health (Fan, Gear, Pump, Impeller etc)
 Shorted Turns
 Cavitations, Unbalance, Misalignment, Foundation Looseness etc
Torque Spectrum
Case Study Bearing Failure

 5hp, 4 pole, Baldor Motor

 24” Axial Fan
 Instrumented with
 Cognitive Systems CV395B Spectrum Analyzer
 Bentley Nevada ADRE 208P (Automated Diagnostics
for Rotating Equipment)
 SWANTECH stress wave analysis system
 Air flow meters
 Accelerometers
 Laser tachometers
 Current meters
 Thermocouples
 Baker Explorer
 SKF 6503 Deep Groove Bearing
 .062 deep, .062 wide grove in outer race.
 Vibration should show at the ball pass
frequency in formula below:(n=#balls,
f=rot freq, Bd=ball dia, Pd= path dia,
beta=contact angle) BPFO=107Hz:

n Bd
Outer Race ( BPFO ) = f (1 − Cosβ )
2 Pd
 3-D Demodulation

n Bd
Outer Race ( BPFO ) = f (1 − Cosβ )
2 Pd
Vibration Spectrum
Known Fault

harm. * BPFO ± 2 * RPM

 Torque Spectrum
BPFO Controlled Lab Test
… how it works
Known Good Bearing Known Outer Race Defect
Electrical Frequencies Removed Adding
Electrical
Electrical
Frequencies
Marking Harmonic
Removed
1 * BPFO Sidebands
Known Good Bearing
Outer Race Fault
120Hz – 240Hz – 120Hz +
BPFO BPFO BPFO BPFO
VFD Details

 Torque and Speed vs. Time

 Frequency and Voltage vs. Time
VFD: Typical issues

 Voltage spikes
 Non-sinusoidal sine wave
 Shaft voltages
 Bearing failures
 Voltage distortion
 Volts/Hertz ratio
Variable Frequency

VFD:
Conveyor Drive.
60 hp, 1200 rpm
Torque

Speed

Frequency

Voltage
 VFD: What is going on?
Conveyor Drive. 60 hp 1200 rpm

• Breaking
• Generator
• Over rated

• Oscillating
• Flawed control loop design
Transient Analysis

Captur es voltage, cur r ent and tor que vs time at star tup
Transient Analysis - Normal start-up

Voltage vs. time Torque vs. time

Current vs. time

Transient Analysis-Case Study

 Using start-up transient analysis to adjust soft-starts and locate problem areas.
 In this case the customer had six new soft-starts installed; five were performing as
expected but one was not.
 Result: New soft-start had to be replaced.
Proper Operation
Problem
Transient analysis: Common signatures
Cavitations

Case Study

 3 Identical Vertical Pumps Feeding a Manifold

 4160V, 1250hp, 254rpm
 Pumps designed to pull 32,000Nm from motor.

 Motor designed to provide 35,000Nm for full load.

 One Pump pulling only 28,000Nm from motor.
 Findings

 Problem found in the pump – not the motor:

 First pump would only run at 28,000Nm or 75% of capacity
 Pump showed fluctuating torque
 The utility wanted higher cooling capacity for summer months - pump
pulled for repair
 Repair cost $180,000
Torque Signature

Pump # 1 Typical - Pump # 2 & 3

The Saga Continues

 The pump was repaired & put back on line.

 Power output of the entire 730MW plant was available…..until
 One of the other pumps sheared its shaft.
 Second Pump Failed

 It has been speculated that the first pump’s turbulent water flow in the manifold overstressed
the shaft of the second pump. The pump may have have been destined for failure regardless
of the first pump’s performance.
 Torque ripple was higher on 2nd and 3rd pump when 1st pump was running at 75% with
loose end bell. Didn’t think too much of this at the time.
 Plant now down to 2/3rd of it’s thermal output because 1/3rd of cooling capacity was lost.
 Repair of the second pump took 5 weeks
 During the 5 weeks, the first pump was able to provide 100% of its rated pumping capacity.
Had it still been at 75% capacity, $4,000,000 in power sales would have been lost due to low
cooling capacity. (E. Wiedenbrug, “Instantaneous Torque as Predictive Maintenance Tool
for Variable Frequency Drives and Line Operated Motors,”)
Induced Imbalance

 7.6 grams introduced at 2” from center of fan – 0.54ounce-inches of

imbalance.
Imbalance Results

 Torque spectra showed a significant difference between the balanced and

unbalanced signal at the 1 x shaft speed. Torque signal very usable in finding
imbalances.
 Accelerometers mounted on the outside of the duct showed a 6.7% increase in
amplitude at 1x shaft speed. Vibration does change, but it would be tough to be
definitive.
Mechanical Problem

Loose Fan Belt Case Study

 Using “Torque Ripple” to locate mechanical issues.
 In this case, an air handler was having bearing failures and alignment problems.
 The torque ripple identified the very cyclic beat of a loose belt.
 Air Handling Fan

 125hp, 480V, 1780RPM motor driving an air

handling fan – 6 ft between pulleys.
 Torque varying 25% +
 Can’t keep motor and fan aligned.
 Replace fan pillow block bearings every 35-
45 days.
 Replace motor bearings every 6 months.
 Use torque deviation to determine when to
realign.
 No longer necessary to replace bearings so
often.